CN211185838U - Tobacco leaf rapid detection grading plant based on image and spectral characteristics - Google Patents

Abstract

The utility model discloses a rapid detection and classification device for tobacco leaves based on image and spectral characteristics, which comprises a frame, a transmission module, a global automatic exposure camera, a near-infrared spectrometer, a control module and a sorting module. The transmission module comprises a conveyor belt, a motor I, and a conveyor belt The conveyor rollers are arranged on the frame and connected with the output shaft of the motor I. The frame is equipped with a global automatic exposure camera and a near-infrared spectrometer in the middle or above the front of the conveyor belt, and the frame is equipped with a sorting module at the rear of the conveyor belt. The global automatic exposure camera, the near-infrared spectrometer and the sorting module are electrically connected to the control module. The global automatic exposure camera and the near-infrared spectrometer transmit the captured images and spectral signals to the control module, and the control module processes the images and spectral signals and then controls the sorting module. The utility model can automatically identify and determine the maturity of fresh tobacco leaves, and can automatically classify and collect, and has the characteristics of accurate classification, high degree of automation, and not easy to damage the tobacco leaves.

Description

A device for rapid detection and classification of tobacco leaves based on image and spectral features

technical field

The utility model belongs to the technical field of tobacco machinery, in particular to a tobacco leaf rapid detection and classification device based on image and spectral characteristics, which is simple in structure, accurate in classification, high in automation, strong in environmental adaptability and not easy to damage tobacco leaves.

Background technique

"The cultivation of tobacco leaves in China dates back to before the Han Dynasty." my country is the world's largest producer of tobacco leaves, and tobacco is an important economic crop in my country. Tobacco farmers pick fresh tobacco leaves, they need to classify the tobacco leaves according to the different parts of the plant, as well as the different maturity, different sizes, water content, starch content and protein content of fresh tobacco leaves, and tie the same grade of tobacco leaves in bundles on the tobacco rods and tobacco. On the rope, the leaves with small leaves or low water content are slightly densely woven, and the leaves with large leaves and high water content are slightly sparsely woven, and then bundles of fresh tobacco leaves of different grades are placed in different positions in the roasting room to prevent the tobacco leaves from being discharged during the roasting process. Rates vary, resulting in brown tobacco leaves. The traditional manual grading of fresh tobacco leaves is cumbersome and can only be completed by tobacco farmers or professionally trained sorting personnel. Manual grading is not only resource-intensive, labor-intensive, and inefficient, but also personnel are easily affected by environmental conditions, emotions, Affected by factors such as the abundance and lack of experience, it is difficult to ensure that the tobacco leaves are sorted to a specific level, and the sorting quality and accuracy are relatively low. In addition, it is easy to cause damage to fresh tobacco leaves during the manual grading process, which will greatly reduce the quality of tobacco leaves, thereby reducing the economic value of tobacco leaves.

With the advancement of machine vision technology, image processing technology can achieve the processing ability to recognize pictures of different sizes and colors. In addition, with the progress of near-infrared technology in plant research, there is also the use of near-infrared offline detection of chemical components in tobacco leaves, and after years of model establishment and maintenance, the deviation between the near-infrared detection value and the actual value is less than 3%, which satisfies the offline detection of enterprises. requirements. Therefore, at present, machine vision or near-infrared technology is combined with automation to achieve automatic grading of tobacco leaves from a certain aspect. However, due to the single evaluation index for tobacco grading, it is difficult to improve the accuracy of grading. There are also automatic combinations of machine vision and near-infrared technology, trying to consider the size, color and chemical composition of tobacco leaves in order to achieve the effect of computer intelligent control, thereby improving the accuracy and quality of tobacco leaf grading. However, in order to ensure the accuracy of machine vision, it is often necessary to set up a special light source and a hood to ensure the stability of the color of the captured image. Generally, an independent photodetector and other auxiliary sensors are also set to determine the position of the tobacco leaves, so as to ensure that the photo is taken. The correct position of the tobacco leaves in the lens is not only complicated in structure and weak in environmental adaptability, but also has a long auxiliary preparation time, a large amount of data, and is easy to damage the tobacco leaves.

Utility model content

The purpose of the utility model is to provide a rapid detection and grading device for tobacco leaves based on image and spectral characteristics, which is simple in structure, accurate in classification, high in automation, strong in environmental adaptability, and not easy to damage tobacco leaves.

The purpose of the utility model is achieved as follows: including a frame, a transmission module, a global automatic exposure camera, a near-infrared spectrometer, a control module, and a sorting module. The frame is connected with the output shaft of the motor I. The frame is fixedly provided with a global automatic exposure camera and a near-infrared spectrometer in the middle or above the front of the conveyor belt feeding direction, and the frame is in the rear of the conveyor belt feeding direction. A sorting module is fixedly arranged, the global automatic exposure camera, the near-infrared spectrometer and the sorting module are respectively electrically connected to the control module, and the global automatic exposure camera and the near-infrared spectrometer transmit the captured images and spectral signals to the control module, The control module controls the sorting module after processing the image and spectral signals.

The beneficial effects of the present utility model:

1. The utility model adopts the global automatic exposure camera to cooperate with the control module, which can automatically detect the entry of tobacco leaves and automatically capture and import the control module. Through the global automatic exposure camera’s ability to capture the details of moving objects, it solves the problem that ordinary cameras cannot obtain moving video content. In addition, the automatic exposure capability of the global automatic exposure camera can automatically adapt to the natural light changes of the application environment, thereby simplifying the shooting auxiliary devices and auxiliary time, and effectively improving the quality and efficiency of shooting images of tobacco leaves.

2. The utility model combines the global automatic exposure camera and the near-infrared spectrometer to collect images and spectral signals from the tobacco leaves moving continuously on the conveyor belt, so that the control module can comprehensively consider the chemical components such as the size, color, starch content, and protein content of the tobacco leaves. , which can accurately determine the grade of tobacco leaves, effectively improve the accuracy and efficiency of tobacco leaf classification, avoid the problem of low accuracy of traditional manual judgment of fresh tobacco leaf maturity, and provide some technical support for the establishment of intelligent tobacco leaf curing.

3. The utility model connects the sorting module with the control module, and the control module can automatically control the sorting module to accurately classify and concentrate the continuously moving tobacco leaves on the conveyor belt according to the grade according to the classification result, avoiding the traditional manual classification of fresh tobacco leaves. The problem of easily damaged tobacco leaves not only improves the sorting quality of tobacco leaves and the integrity of tobacco leaves, but also reduces the labor intensity in the process of grading and collecting tobacco leaves.

Description of drawings

Fig. 1 is the structural principle schematic diagram of the present utility model;

Fig. 2 is a schematic diagram of the structure principle of the sorting module of Fig. 1;

In the picture: 1-rack, 2-conveyor module, 201-conveyor belt, 3-global automatic exposure camera, 4-near infrared spectrometer, 5-control module, 6-sorting module, 601-chain, 602-sprocket, 603-brush, 7-light source, 8-light baffle, 9-collection frame.

Detailed ways

The present utility model is further described below in conjunction with the accompanying drawings and examples, but the present utility model is not limited in any way, and any changes or improvements made based on the teachings of the present utility model belong to the protection scope of the present utility model.

As shown in FIG. 1 and FIG. 2 , the present invention includes a frame 1, a transmission module 2, a global automatic exposure camera 3, a near-infrared spectrometer 4, a control module 5, and a sorting module 6. The transmission module 2 includes a conveyor belt 201, Motor I, the conveyor belt 201 is installed on the frame 1 through the conveying rollers and connected to the output shaft of the motor I. The frame 1 is fixedly provided with a global automatic exposure camera 3 and a close-up camera in the middle or above the front of the conveyor belt 201 in the incoming direction. Infrared spectrometer 4, the rack 1 is fixedly provided with a sorting module 6 at the rear of the conveyor belt 201 in the incoming direction, the global automatic exposure camera 3, the near-infrared spectrometer 4 and the sorting module 6 are respectively electrically connected with the control module 5 , the global automatic exposure camera 3 and the near-infrared spectrometer 4 transmit the captured images and spectral signals to the control module 5, and the control module 5 controls the sorting module 6 after processing the images and the spectral signals.

The camera head of the global automatic exposure camera 3 and/or the probe head of the near-infrared spectrometer 4 are perpendicular to the belt surface of the conveyor belt 201 .

The control module 5 includes a processor, a memory, and an I/O module. The processor is connected to the signal ports of the global automatic exposure camera 3, the near-infrared spectrometer 4 and the sorting module 6 through the I/O module. A tobacco leaf algorithm model is stored, and the processor is used to import the image captured by the automatic exposure camera 3 and the spectral signal of the near-infrared spectrometer 4 into the tobacco leaf algorithm model, and then control the sorting module 6 according to the calculation result.

A light source 7 is arranged around the camera of the global automatic exposure camera 3 .

Below the camera side of the global automatic exposure camera 3, a standard whiteboard for correcting the white balance is fixedly arranged on the frame 1 on at least one side of the conveyor belt 201, which is perpendicular to the camera.

A light shielding plate 8 is fixed on the frame between the global automatic exposure camera 3 and the near-infrared spectrometer 4 .

The sorting module 6 is provided with a plurality of collecting frames 9 for receiving tobacco leaves under one side or both sides of the conveyor belt 201 .

The sorting module 6 includes an air blowing pipe, an air pump, and a nozzle whose outlet direction is perpendicular to the conveyor belt 201, which are fixedly arranged on one or both sides of the conveyor belt 201. The air blowing pipe is communicated with the air pump, and the air blowing pipe Nozzles are provided along the conveyor belt 201 corresponding to the collecting frame 9 .

The air blowing pipe and the air pump are connected in series with a solenoid valve electrically connected to the control module 5 . The control module 5 controls the opening and closing of the solenoid valve to control the nozzle to blow air so that the tobacco leaves on the conveyor belt 201 fall into the collection frame 9 .

The sorting module 6 includes a plurality of motors II, a chain 601, a sprocket 602, and a brush 603 that are sequentially distributed along the incoming direction of the conveyor belt 201. The motor II 601 is fixedly arranged on the frame 1 and is controlled by the sorting module 6. The output shaft of the motor II 601 is connected to the chain 601 , the chain 601 is perpendicular to the conveyor belt 201 and is sleeved on the sprocket 602 , and one end of the brush 603 is fixed to the end face of the chain 601 .

The working principle and working process of the utility model:

The utility model adopts the global automatic exposure camera to cooperate with the control module, which can automatically detect the entry of tobacco leaves and automatically capture and import into the control module. It can automatically adapt to the natural illumination changes of the application environment by using the automatic exposure capability of the global automatic exposure camera, thereby simplifying the shooting auxiliary device and auxiliary time, and effectively improving the shooting image quality and shooting efficiency of tobacco leaves. The global automatic exposure camera and near-infrared spectrometer collect images and spectral signals from the continuously moving tobacco leaves on the conveyor belt, so that the control module can accurately identify the tobacco leaves from the size, color, starch content, protein content and other chemical components of the tobacco leaves. It can effectively improve the accuracy and efficiency of tobacco leaf classification, avoid the problem of low accuracy of traditional manual judgment of fresh tobacco leaf maturity, and provide partial technical support for the establishment of intelligent tobacco leaf curing; the utility model integrates a sorting module and a control module. Connected, the control module can automatically control the sorting module to accurately classify and concentrate the continuously moving tobacco leaves on the conveyor belt according to the grading results, which avoids the traditional manual grading of fresh tobacco leaves. The quality of the sorting and the integrity of the tobacco leaves are improved, and the labor intensity during the grading and collection of the tobacco leaves is also reduced. Further, the camera of the global automatic exposure camera and/or the probe of the near-infrared spectrometer are perpendicular to the belt surface of the conveyor belt, which can not only ensure that the image is not deformed, avoid post-correction, improve the processing speed, and reduce the effects of ambient light and background changes. Therefore, the accuracy of the collected data can be improved. Furthermore, the light source is arranged around the camera of the global automatic exposure camera, which can fill in the light when the ambient illumination is low, so that the utility model can adapt to the indoor environment at night and with insufficient illumination, and improve the adaptability of the environment. Further, a standard whiteboard for correcting white balance is set under the camera of the global auto exposure camera, which can automatically perform white balance correction before the camera is started and before the tobacco leaf image is captured, thereby eliminating the influence of ambient light changes and improving the color accuracy of the tobacco leaf image. . Further, the sorting module can use the electromechanical chain driven brush rotation and the air blowing structure to collect the tobacco leaves on the conveyor belt in stages according to the needs and conditions, which can improve the sorting efficiency and reduce the loss of tobacco leaves. . To sum up, the utility model has the characteristics of simple structure, accurate classification, high degree of automation, strong environmental adaptability, and not easy to damage tobacco leaves.

As shown in Figures 1 and 2, start the equipment, the motor I in the transmission module 2 drives the conveyor belt 201 to run, and at the same time the global automatic exposure camera 3, the near-infrared spectrometer 4, the sorting module 6 and the computer 5 (control module) self-check, After the self-inspection is passed, the computer 5 controls the global automatic exposure camera 3 or the global automatic exposure camera 3 to automatically shoot the image of the standard whiteboard 7, and compares the white balance of the standard whiteboard 7 in the identification image with the preset actual white balance value to complete the white balance. Correction.

After the smokers collect the tobacco leaves, they neatly spread the tobacco leaves on the conveyor belt 201 with a moving speed of 0.1m/s, and the interval between each tobacco leaf is about 0.15m. , the captured image is sent to the computer 5, the tobacco leaf algorithm model established on the AI open platform in the computer 5 automatically detects the entry of the tobacco leaf, controls the global automatic exposure camera 3 to capture the tobacco leaf panoramic image and imports it, and then determines the tobacco leaf color and color according to the tobacco leaf algorithm model. size. The tobacco leaves continue to move with the conveyor belt 201 to the probe of the near-infrared spectrometer 4 , and the chemical components including starch content and protein content of the fresh tobacco leaves are measured and sent to the computer 5 . The computer 5 automatically analyzes the data in the database according to the received tobacco leaf images and infrared spectrum signals and makes a grade evaluation of fresh tobacco leaves. Subsequently, the computer 5 controls the action of the sorting module 6 to divert the tobacco leaves into the collection baskets 8 of the corresponding tobacco leaf maturity levels.

Claims (10)

1. A device for rapid detection and classification of tobacco leaves based on image and spectral characteristics, characterized in that it comprises a frame (1), a transmission module (2), a global automatic exposure camera (3), a near-infrared spectrometer (4), a control module ( 5) A sorting module (6), the conveying module (2) includes a conveying belt (201) and a motor I, the conveying belt (201) is arranged on the frame (1) through conveying rollers and is connected to the output shaft of the motor I , a global automatic exposure camera (3) and a near-infrared spectrometer (4) are fixedly installed in the frame (1) in the middle or above the front of the conveyor belt (201) in the incoming material direction, and the frame (1) is on the conveyor belt (201) ) a sorting module (6) is fixedly arranged at the rear of the incoming material direction, the global automatic exposure camera (3), the near-infrared spectrometer (4) and the sorting module (6) are respectively electrically connected with the control module (5), The global automatic exposure camera (3) and the near-infrared spectrometer (4) transmit the captured image and the spectral signal to the control module (5), and the control module (5) controls the sorting module (6) after processing the image and the spectral signal ). 2. The device for rapid detection and grading of tobacco leaves based on image and spectral features according to claim 1, characterized in that the camera head of the global automatic exposure camera (3) and/or the probe of the near-infrared spectrometer (4) are perpendicular to the conveyor belt (201). ) of the belt surface. 3. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to claim 2, wherein the control module (5) comprises a processor, a memory, and an I/O module, and the processor passes through the I/O module. is connected to the signal port of the global automatic exposure camera (3), the near-infrared spectrometer (4) and the sorting module (6), the memory stores a tobacco leaf algorithm model, and the processor is used for shooting the automatic exposure camera (3) The image and the spectral signal of the near-infrared spectrometer (4) are imported into the tobacco leaf algorithm model, and then the sorting module (6) is controlled according to the operation result. 4. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to claim 2, characterized in that a light source (7) is provided around the camera of the global automatic exposure camera (3). 5. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to claim 1, characterized in that the frame (1) on at least one side of the conveyor belt (201) is located below the camera side of the global automatic exposure camera (3). A standard whiteboard is fixed on the top to correct the white balance perpendicular to the camera. 6. The device for rapid detection and grading of tobacco leaves based on image and spectral characteristics according to claim 1, characterized in that a light baffle is fixed on the frame between the global automatic exposure camera (3) and the near-infrared spectrometer (4). (8). 7. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to any one of claims 1 to 6, characterized in that the sorting module (6) is provided with a conveyor belt (201) under one side or both sides. A plurality of collecting boxes (9) for receiving tobacco leaves. 8. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to claim 6, characterized in that the sorting module (6) comprises a rack (1) fixedly arranged on one or both sides of the conveyor belt (201). The air blowing pipe, the air pump, and the nozzle whose outlet direction is perpendicular to the conveyor belt (201), the air blowing pipe is communicated with the air pump, and the air blowing pipe is provided along the conveyor belt (201) corresponding to the collecting frame (9). 9. The device for rapid detection and classification of tobacco leaves based on image and spectral characteristics according to claim 8, characterized in that the air blowing pipe and the air pump are connected in series with a solenoid valve electrically connected to a control module (5), and the control module (5) ) controls the opening and closing of the solenoid valve to control the blowing of the nozzle so that the tobacco leaves on the conveyor belt (201) fall into the collecting frame (9). 10. The device for rapid detection and classification of tobacco leaves based on image and spectral features according to claim 6, characterized in that the sorting module (6) comprises a plurality of motors II, chains ( 601), sprocket (602), brush (603), the motor II is fixedly arranged on the frame (1) and is electrically connected with the controller of the sorting module (6), the output shaft of the motor II is connected to the A chain (601) is connected, the chain (601) is perpendicular to the conveyor belt (201) and is sleeved on the sprocket (602), and one end of the brush (603) is fixed to the end face of the chain (601).

Images